To Jeroen and interested minds, re. Tiger node count

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
Gerd Isenberg
Posts: 2233
Joined: Wed Mar 08, 2006 7:47 pm
Location: Hattingen, Germany

Re: To Jeroen and interested minds, re. Tiger node count

Post by Gerd Isenberg » Sat Aug 16, 2008 1:57 pm

Jeroen wrote:http://www.mijnwoordenboek.nl/vertalen.php


collaborator(the ~) als in `traitor to one's country`:
landverrader (de ~ (m))
collaborator(the ~) als in `quisling`:
landverrader (de ~ (m))
collaborator(the ~) als in `collaborator`:
collaborateur (de ~ (m)), landverrader (de ~ (m))

When you are in Holland again, use the word a few times addressing Dutch people and see what happens.
Hi Jeroen,

I was formerly not aware about the disambiguation of collaboration or collaborator, but it seems quite usual in English speaking countries for people (for instance chess programmers) working together or cooperate, versus working with an outside entity against his own society or faction.

Similar disambiguation appears to Euthanasia in German speaking countries, it has stigma of killing handicapped people in Nazi-Germany.

By the way, congrats to the whole Rybka team for that outstanding success and all the best and commercial success. At least we all others are aware there is still some progress possible!

Gerd

User avatar
Eelco de Groot
Posts: 4279
Joined: Sun Mar 12, 2006 1:40 am
Location: Groningen

Re: To Jeroen and interested minds, re. Tiger node count

Post by Eelco de Groot » Sat Aug 16, 2008 2:38 pm

Jeroen wrote:http://www.mijnwoordenboek.nl/vertalen.php


collaborator(the ~) als in `traitor to one's country`:
landverrader (de ~ (m))
collaborator(the ~) als in `quisling`:
landverrader (de ~ (m))
collaborator(the ~) als in `collaborator`:
collaborateur (de ~ (m)), landverrader (de ~ (m))

When you are in Holland again, use the word a few times addressing Dutch people and see what happens.
Before Harvey starts using this as a newly learnt Dutch insult in Leiden, my Van Dale dictionary gives a more strict definition:

collaborator (Lat.), m. (-s), medewerker

which is what it means in Latin and is the usual meaning in English too.

The negative association from WW II, of people helping the occupying German forces against their own compatriots is to my knowledge more with the French word 'collaborateur' What exactly the origin is of the use of this French idiom in this context I do not know.

Eelco

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Numbers have no absolute meaning

Post by bob » Sat Aug 16, 2008 3:12 pm

Rolf wrote:
bob wrote:
Rolf wrote:
kranium wrote:Chris W. is absolutely right here...and has expressed it perfectly.
I know Rolf is pressuring you, but you owe no explanantion.
I was convinced Jeroen had a good question. What I expected that either CT admitted or rejected that he also gave wrong numbers in display. That was only interesting in a relational sense. IMO all programmers can do what they want with their output. Only Bob, CT and Norm saw it different. What is correct asks the lay...
Let's set the record straight. I never gave diddly-squat about Rybka's node count, obfuscated or not. I didn't discover the obfuscation. I didn't measure the obfuscation. I didn't verify the obfuscation. None of it mattered to me. So let's get that crap out of the sandbox before we play any more.

Someone asked a simple question "why would one do this?" and I gave a very specific and precise answer I didn't make any accusations. I didn't ask any questions. I didn't run any tests. I just explained _why_ one would want to obfuscate specific output of their engine, nothing more, nothing less. I do have the opinion that if one claims that they searched XXX nodes, then they ought to use the _standard_ definition of "node". Even if there is room to count things more than one way. Because this "room" is still a very small percentage of total nodes.

My only interest is in vocabulary and definitions. Everyone counts a ply the same way, with one known exception. Depth is a bit more variable since you could report it as minum full-width depth, average depth over all branches, or even as a multiple-parameter value like 10/15/30 for min, average and max. Nodes are precisely defined, and not just for computer chess, but for _all_ tree-searching applications including checkers, go, and other problems like the travelling salesman example. If we don't use a common vocabulary, with standard and accepted definitions, then we have problems communicating. Much like the spacecraft launched to either mars or some moon out that way, which landed 10 meters below the surface of the planet because one group was measuring in meters, the other was using feet. The way to avoid that is to use absolutely standard definitions. In math, everyone knows what "floor" means, and "ceil", and "1st derivative" We need the same preciseness here.

hope that clears up where _I_ stand on the issue.
Some days ago I wrote something about military and secret services and you took that as a an absolute nonsense but I had something concrete in mind. And here I can work it in again.

I still think that in military research, academic research with military impact, research in highest interest of State raison, and for example computerchess you cant expect to reveil secrets of actual standings of the research. Just as examples. The military will never exactly tell you the absolute distance of weapons, of photographic resolution from the Space, of the existence of lethal weapons as such. In computerchess competition someone would be a fool if he gave away exact data.

Of course the tournament organisers could order a sort of doping test for all winners, but this would probably mean few participants because the revelation of exact secrets dont make sense for commercial entities.

In times of the leadership of SMK it was similar. In average tournaments Shredder could lose but in Championships with Stefan as operator Shredder won again. Reason: for such tournaments special - non-public, non-tested versions were in use.

Vas has some other tricks and all others have still others. No chance for reveiling standards for the eval output. And it also wouldnt make sense IMO because different versions of the same engine show a completely different eval score. Just look at Rybka 3 and then its sisters R3 Human and R3 Dynamic. MAnsari reported first results that Dynamic has better results in long time controld. For the same positions the different variants show very different evals! So I would expect that the user should never take numbers as sort of exact copies of human chess game commentary in Informator style and classical numbers for different standings.
All well and good, but it misses the point. A simple example. Why do most cars have tachometers in them? Does anyone really use that to shift since most are now using automatic transmissions? Do you drive down the highway at 3,000 RPM or at XXX MPH/KPH??? So is there any reason why a manufacturer should not obfuscate the tach in your car and make it show 1/10th the RPM?

that is what the discussion is about here. There are plenty of cars made without a tach. But those that do have one are displaying the _same_ data so that you can compare the two vehicles. Is it important to be able to do so? I'd rather drive an auto on the interstate at 2,000 RPM than drive one that has to turn 7,000 to reach that speed. the 2K motor is going to last 10x longer. And the minute you give me that tach with accurate numbers, I can now discover exactly what the transmission gear ratios are, what the final drive ratio is, where the transmission shifts under various driving conditions, what the torque curve of the motor looks like across the rpm band. Etc. I can discover that there is a strong surge in power at 4,500 rpm, and that you are using a variable-length-manifold-runner mechanism to change the intake runner length at 4500 RPM. In short, I can discover a +lot+ about the overall behavior of the car, the fact that it doesn't have any gears and is using a continuously-variable-ratio drive system rather than a traditional N-speed transmission. If you want to hide that better, don't give me the RPM. Or give me an RPM that is outright false. except that by doing the latter, I might well catch on that your data is false, while without presenting any data at all, I would have no idea.

So why present _false_ data, and then get caught doing so and have to explain why, when we already know why?

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: To Jeroen and interested minds, re. Tiger node count

Post by bob » Sat Aug 16, 2008 3:25 pm

Jeroen wrote:Huh? You must be kidding. Calling me a 'collaborator' (which is a very negative word in the Dutch language) is pretty offensive, so I am filing an official complaint here. I don't accept such language, especially not from a moderator.

So basically other programmers can question anything about Rybka and not vice versa. OK, you made that perfectly clear.

In that case I wish you a lot of success with the 'impartial CCC where everybody can ask what he wants'. Maybe you should change its name into 'The Anti Rybka Forum'.
While I didn't read the post in question, "collaborator" is hardly an insulting word. It is what I call someone that works on something with me. IE we "collaborate" to improve something... hardly an insult, just based on the word... More like "friend helping me" or something similar. For example, in the thread on thesting on the programming forum, I would call Karl a collaborator and hope he would not be insulted...

Jeroen
Posts: 501
Joined: Wed Mar 08, 2006 8:49 pm

Re: To Jeroen and interested minds, re. Tiger node count

Post by Jeroen » Sat Aug 16, 2008 3:50 pm

Hi Gerd,

Yes, words sometimes can have different meanings in other languages. I remember the discussion about 'euthanasie', which is legal in Holland under strict rules, but has a different meaning in other countries.

Anyway, I think I overreacted and Chris has stated that he meant 'working together with'. For me the first thing reading the word 'collaborator' is a connection with the meaning 'traitor', which many of my countrymen will also do.

So apologies to Chris and the other CCC readers for misinterpreting what Chris was telling me.

Jeroen

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 10:14 pm
Location: Munster, Nuremberg, Princeton

Re: To Jeroen and interested minds, re. Tiger node count

Post by Rolf » Sat Aug 16, 2008 4:19 pm

bob wrote:
Jeroen wrote:Huh? You must be kidding. Calling me a 'collaborator' (which is a very negative word in the Dutch language) is pretty offensive, so I am filing an official complaint here. I don't accept such language, especially not from a moderator.

So basically other programmers can question anything about Rybka and not vice versa. OK, you made that perfectly clear.

In that case I wish you a lot of success with the 'impartial CCC where everybody can ask what he wants'. Maybe you should change its name into 'The Anti Rybka Forum'.
While I didn't read the post in question, "collaborator" is hardly an insulting word. It is what I call someone that works on something with me. IE we "collaborate" to improve something... hardly an insult, just based on the word... More like "friend helping me" or something similar. For example, in the thread on thesting on the programming forum, I would call Karl a collaborator and hope he would not be insulted...
Bob, allow me please if I disagree.

JN and CT once were partners in Tiger buisiness and competition. Believe me Bob, CT had well learned this specific Dutch historical problem connected with the Nazi occupation. How ugly the term is for Dutch, all could see when I used it a bit careless in a debate with Ed longtime ago. He went almost mad at me imediately.

We have here a typical hypocrisy problem. Someone does something intentionally to push the buttons of another one whose national weaknesses here and NOT at all personal ones he well knows and the many experts here, also two mods, also CW, and now you explain Jeroen that this wasnt insultive at all, as if they wanted to humiliate him another time! It's really ugly.

It's utmost impolite if you intentionally leave out all psychology of such a topic and then inform a perfectly innocent here Dutch that he has made his own problem now because he simply didnt get the absolute innocence of an expression!

In real it's completely different. CT made an intentional insult. After he made it against Vas he now made it against Jeroen and out of his view all this is justified and this is the "crime".

In truth his whole campaign against Vas was already the "crime". And because JN now dared to show CT his own practice, CT is now running out of control. Bob, he already announced it, "It's not over yet" he told Jeroen. But you and others still think that this is a misunderstanding?

After targeting Vas and Jeroen, the next big topic will be ChessBase for sure. It will start a campaign against the true kingdom of evil and the bloodsucking, when in truth there is nothing but a repetition of a well known trend, namely thatr all best engines once in a while had a date with CB and appeared under the Fritz GUI or their own. In the meantime you can download and buy some dozen programs on the CB webpage, among those are Junior, Comet and Crafty. And now also Rybka. And formerly Tiger and also Sjeng. You name it. But there will be a defamatory campaign against CB and Rybka, because their connection. And it all started this summer with CT's campaign against Vas and his KN count.

What I want to say to you in special is this, Bob. You are the guard for state of the art in computerchess at least who is still active on the net. You care for real KN count. Fine. But then if you are so active, then please give your clear commentary, that CB or Vas with Rybka are no criminals who have violated the old charme of good old ACM laws. We have 2008 and some people want to make a living out of the pretty small business niche that computercxhess has to offer. It were you who should explain and not me, that some plays are still allowed and that this here is not university. I mean when it comes to the idiotic Elo numbers and other stats you also show tolerance for such games although you know that science wise it's nonsense.

Therefore I would like that you as our number one top reference wouldnt tell Jeroen to shut up with his being hurt, and that you would tell CT that he should better stop his comedy smear act against Vas. That he should focus on Tiger as Theron, because what has Theron to do with Rybka suddenly??? I wished you would make such clarifications.

Tip to hat regards, Bob. Nice weekend for you.
-Popper and Lakatos are good but I'm stuck on Leibowitz

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 10:14 pm
Location: Munster, Nuremberg, Princeton

Re: Numbers have no absolute meaning

Post by Rolf » Sat Aug 16, 2008 4:51 pm

bob wrote:
All well and good, but it misses the point. A simple example. Why do most cars have tachometers in them? Does anyone really use that to shift since most are now using automatic transmissions? Do you drive down the highway at 3,000 RPM or at XXX MPH/KPH??? So is there any reason why a manufacturer should not obfuscate the tach in your car and make it show 1/10th the RPM?

that is what the discussion is about here. There are plenty of cars made without a tach. But those that do have one are displaying the _same_ data so that you can compare the two vehicles. Is it important to be able to do so? I'd rather drive an auto on the interstate at 2,000 RPM than drive one that has to turn 7,000 to reach that speed. the 2K motor is going to last 10x longer. And the minute you give me that tach with accurate numbers, I can now discover exactly what the transmission gear ratios are, what the final drive ratio is, where the transmission shifts under various driving conditions, what the torque curve of the motor looks like across the rpm band. Etc. I can discover that there is a strong surge in power at 4,500 rpm, and that you are using a variable-length-manifold-runner mechanism to change the intake runner length at 4500 RPM. In short, I can discover a +lot+ about the overall behavior of the car, the fact that it doesn't have any gears and is using a continuously-variable-ratio drive system rather than a traditional N-speed transmission. If you want to hide that better, don't give me the RPM. Or give me an RPM that is outright false. except that by doing the latter, I might well catch on that your data is false, while without presenting any data at all, I would have no idea.

So why present _false_ data, and then get caught doing so and have to explain why, when we already know why?
Bob, I take it for sure that you are seated because this will knock you out of your shoes. Yes, this is all correct what you write about cars, and the tacho is enormously important for - well even for James Dean it was. But Vas, hey, I didnt ask him, but I know that he has explained what his numbers mean and that numbers in comparison between different competitors is nonsense. But I know something that will be published here for the first time.

Vas is a chessplayer himself and he wants to inform his users that he is interested in strong chess and not tacho numbers. High above in the thread I elaborated what I thought about a display of +14. for a chess where White has a free Pawn on b5 already. Please take a look and please make a short remark to my idea. Because these +14. which traditionally means that White has already won three games in a row, is something that a machine could be told from a chessplayer. As a potential that then must be focused and then transfered to the KO. Of course with Kramnik as Black the +14. are idiotic.

I want to publish that for the first time in history we have a programmer who wants to make his baby play good chess, and not just good stats in the test tournament. A program that finds the Anand move Nc7! Qxc7 Rc8 Qxe7 Nc4 Bc5 Rxa3 etc.!!!

To me the reactions here about the KN count in Rybka is the similar reaction CSTal had to overcome. It's the displayed arrogance of seemingly mathematically correct programmers who insinuate that chess is just an engineering process and not also an art. A fight.

And therefore to all my call: Hail Rybka, hail Vas, hail chess and no to all beancounting lifetime wasting.

Gimmi 5! Peace!
-Popper and Lakatos are good but I'm stuck on Leibowitz

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Numbers have no absolute meaning

Post by bob » Sat Aug 16, 2008 5:54 pm

Rolf wrote:
bob wrote:
All well and good, but it misses the point. A simple example. Why do most cars have tachometers in them? Does anyone really use that to shift since most are now using automatic transmissions? Do you drive down the highway at 3,000 RPM or at XXX MPH/KPH??? So is there any reason why a manufacturer should not obfuscate the tach in your car and make it show 1/10th the RPM?

that is what the discussion is about here. There are plenty of cars made without a tach. But those that do have one are displaying the _same_ data so that you can compare the two vehicles. Is it important to be able to do so? I'd rather drive an auto on the interstate at 2,000 RPM than drive one that has to turn 7,000 to reach that speed. the 2K motor is going to last 10x longer. And the minute you give me that tach with accurate numbers, I can now discover exactly what the transmission gear ratios are, what the final drive ratio is, where the transmission shifts under various driving conditions, what the torque curve of the motor looks like across the rpm band. Etc. I can discover that there is a strong surge in power at 4,500 rpm, and that you are using a variable-length-manifold-runner mechanism to change the intake runner length at 4500 RPM. In short, I can discover a +lot+ about the overall behavior of the car, the fact that it doesn't have any gears and is using a continuously-variable-ratio drive system rather than a traditional N-speed transmission. If you want to hide that better, don't give me the RPM. Or give me an RPM that is outright false. except that by doing the latter, I might well catch on that your data is false, while without presenting any data at all, I would have no idea.

So why present _false_ data, and then get caught doing so and have to explain why, when we already know why?
Bob, I take it for sure that you are seated because this will knock you out of your shoes. Yes, this is all correct what you write about cars, and the tacho is enormously important for - well even for James Dean it was. But Vas, hey, I didnt ask him, but I know that he has explained what his numbers mean and that numbers in comparison between different competitors is nonsense. But I know something that will be published here for the first time.

Vas is a chessplayer himself and he wants to inform his users that he is interested in strong chess and not tacho numbers. High above in the thread I elaborated what I thought about a display of +14. for a chess where White has a free Pawn on b5 already. Please take a look and please make a short remark to my idea. Because these +14. which traditionally means that White has already won three games in a row, is something that a machine could be told from a chessplayer. As a potential that then must be focused and then transfered to the KO. Of course with Kramnik as Black the +14. are idiotic.

I want to publish that for the first time in history we have a programmer who wants to make his baby play good chess, and not just good stats in the test tournament. A program that finds the Anand move Nc7! Qxc7 Rc8 Qxe7 Nc4 Bc5 Rxa3 etc.!!!

To me the reactions here about the KN count in Rybka is the similar reaction CSTal had to overcome. It's the displayed arrogance of seemingly mathematically correct programmers who insinuate that chess is just an engineering process and not also an art. A fight.

And therefore to all my call: Hail Rybka, hail Vas, hail chess and no to all beancounting lifetime wasting.

Gimmi 5! Peace!
Does anyone claim that +3.0 from program A is equivalent to +3.0 from program B? No more than we consider a +/- by one player to be equal to a judgement of +/- by another player. Those are _opinions_ until an actual win can be demonstrated. I can point out _dozens_ of games in a book like MCO10 or MCO11 where Evans had a final evaluation of +/- and yet black has a crushing move that wins outright. I've been on the bad end of a couple of those in past ACM events.

So evaluations are subjective to say the least. Nodes and such are not. I don't care what he writes, how he tries to explain what he did, it is all pure bullshit. Just as a car displaying the wrong RPM would be pure bullshit. But nobody cares, you say. _some_ do. Particularly those with enough technical savvy to actually extract useful information from that tach. If my overdrive goes out, I will notice it instantly because I will notice the increased RPM at the same speed. You, perhaps, might blindly drive on, increasing the wear rate on your engine by 25% or more (strain increases with square of RPM). So, some care about accurate data. Some could just have multi-colored lights that flash at random times.

You can have whatever you want. But in this case I _know_ what I want, and it isn't ambiguous, and it isn't going to be explained away by someone struggling to justify something after getting caught red-handed. It's as simple as that, IMHO. It is both amusing and a waste of time to try to justify something that can't be justified.

No matter how hard the syncopants continue to try.

User avatar
tiger
Posts: 819
Joined: Sat Mar 11, 2006 2:15 am
Location: Guadeloupe (french caribbean island)
Contact:

Re: To Jeroen and interested minds, re. Tiger node count

Post by tiger » Sat Aug 16, 2008 5:54 pm

Jeroen wrote:Hi Christophe,

Please check your PM's. Let's get this straight as 2 grown ups. I am afraid we are both behaving a little bit like little kids.

Thanks, Jeroen


I have no problem to settle this in the open, in this public forum.

So there is no need to use private messages.

We do not want to leave the impression that either you or I have something to hide, do we?



// Christophe

User avatar
mclane
Posts: 18389
Joined: Thu Mar 09, 2006 5:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub
Contact:

Re: To Jeroen and interested minds, re. Tiger node count

Post by mclane » Sat Aug 16, 2008 6:01 pm

jeroen. words have many meanings.
collaborator has many meanings. it depends which meaning someone wanted to express.

Post Reply